PREFERENTIAL RULES OF ORIGIN POLICY RESEARCH REPORT TRADE ISSUES IN. EAST ASIA June 2007 THE WORLD BANK

Transcription

1 Public Disclosure Authorized TRADE ISSUES IN EAST ASIA June 2007 Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized PREFERENTIAL RULES OF ORIGIN POLICY RESEARCH REPORT EAST ASIA AND PACIFIC REGION POVERTY REDUCTION AND ECONOMIC MANAGEMENT THE WORLD BANK

2

3 Acknowledgments This report was edited by Mona Haddad (Task Team Leader, EASPR). The authors include Titik Anas (CSIS Indonesia), John Lawrence Avila (University of Asia and the Pacific, Philippines), Olivier Cadot (University of Lausanne, CERDI and CEPR), Kakada Dourng (Economic Institute of Cambodia), Robert Kirk (The Services Group Inc., USA), Miriam Manchin (Centro Studi Luca d Agliano, Milan), George Manzano (University of Asia and the Pacific, Philippines), Jaime de Melo (University of Geneva, CERDI and CEPR), Annette Pelkmans-Balaoing (Erasmus University Rotterdam), Alberto Perez-Portugal (University of Geneva), Thiphaphone Phetmany (Enterprise and Development Consultants Co. Ltd, Laos), Apichart Prasert (Fiscal Policy Research Institute, Thailand), Luz Julieta Rio (Enterprise and Development Consultants Co. Ltd, Laos), Hach Sok (Economic Institute of Cambodia), Dang Nhu Van (Institute of Economics, Vietnam), Yang Zerui (CNCPEC, CIIS, China). Special thanks go to Paul Brenton (PRMTR) for his guidance throughout this report and to Silja Baller (EASPR) for putting the report together. The views expressed in this report are the authors own and do not necessarily represent those of the World Bank or the authors respective institutions, or those of the Executive Directors of the World Bank or the governments they represent.

4 Table of Contents Overview i PART I: RULES OF ORIGIN IN EAST-ASIA AND LESSONS FROM OTHER REGIONS 1. Rules of Origin in East Asia 3 1. Preferential Trade Agreements and the Pattern of Trade in East Asia 3 2. AFTA Utilization of Preferences 6 3. Preferential Trade Agreements and Rules of Origin 7 4. The ASEAN Rules of Origin Comparing ASEAN ROO with other FTA ROO Preferential ROO and Regional Integration in East Asia The Way Forward Rules of Origin and the Web of FTAs Introduction Lessons from Existing Preferential Trade Agreements Rules of Origin in East Asia and South-East Asia 28 i. Overview of the Main Rules of Origin provisions in East Asian FTAs 28 ii. Problems Surrounding the Use and Implementation of ROO The Importance of Preferences in AFTA Conclusion Rules of Origin for Preferential Trading Arrangements: Implications for AFTA of EU and US Regimes Introduction Rules of Origin in the US and EU PTAs How Do ROO Affect the Utilization of Preferences Decomposition of Costs Associated with ROO Value Content Rules and Preference Utilization: Estimates from EU Preferential Market Schemes Lessons from NAFTA and EU FTAs 54 PART II: COUNTRY STUDIES: RULES OF ORIGIN FACED BY EAST-ASIAN EXPORTERS 4. Cambodia Export Profile Cambodia s Preferential Agreements Preferential Agreements and Rules of Origin Use of Preferences within ASEAN and China Conclusion and Recommendations China China-ASEAN Economic Relationship China s Free Trade Agreements Rules of Origin in China s Free Trade Agreements ACFTA and its Rules of Origin Its First Two Years Indonesia Trade Profile of ASEAN and Indonesia Preferential Regional Agreements and Rules of Origin Utilization of Form D under AFTA CEPT 90

5 7. Lao PDR Overall Economic Environment Trade Liberalization External Trade Performance Market Access in Specific Sectors 97 i. Textiles 97 ii. Wood processing 98 iii. Agro-processing 99 iv. Cement manufacturing Preferential Agreements and Rules of Origin Survey Results Summary Statements from Respondents Philippines Export Profile Preferential Agreements Case Studies 110 i. Electronics 110 ii. Garments and textiles 113 iii. Motor vehicle parts and components Conclusion Thailand Export Profile Thailand s Free Trade Agreements and Rules of Origin What Discourages the Use of Preferences? Vietnam Export Profile Preferential Agreements and Rules of Origin Findings from Surveys and Interviews 133

6

7 Overview Rules of Origin in East Asia: How Are They Working in Practice? Mona Haddad, EASPR, World Bank East Asia has moved from the rampant unilateralism of the late 1980s, during which external tariffs declined substantially, to the rampant regionalism of the 2000s that is creating a complex Noodle Bowl of preferential agreements (Baldwin (2005)). Until recently, regionalism in East Asia was limited to the ASEAN Free Trade Area (AFTA) which was established in 1992 tariff reductions were fully implemented by the six major ASEAN countries in 2003 while the lowerincome ASEAN members received a longer time frame for implementation. Since 2000, however, new regional agreements have surged in East Asia. The ASEAN-China FTA was signed in Negotiations are also ongoing between ASEAN and several countries including Korea and Japan. The large countries of ASEAN mainly Thailand, Singapore, Indonesia, Malaysia are also contemplating or negotiating several FTAs (with US, Australia, Japan, and others). The increasing number of FTAs in East Asia and elsewhere is creating a complex web of multiple preferential tariff rates applied to various trading partners, often with varying timelines for reducing tariffs. It is also creating a complex and inconsistent web of rules of origin that often limit the use of the trade preferences. Rules of origin are a necessary feature of any regional trade agreement. They ensure that preferences are available only to the signatories of the agreement and that imports from nonmembers do not avoid customs duties by entering through the member with the lowest tariff. The rules of origin define the amount of local processing, or the extent of the transformation of the product, that must be undertaken in the country from which the product claiming preferences is exported. The definition of these requirements to prevent trade deflection is not straightforward. If the rules are too onerous and complex and are costly to comply with, they will limit the impact of tariff reductions on trade. Indeed, import competing industries have often been successful in obtaining restrictive rules of origin that dilute the impact of the loss of tariff protection. This is most apparent in agreements between developed countries and lower wage countries. As FTAs multiply in the region, putting in place rules of origin (ROO) that are simple, transparent, and easy to implement becomes important. The experience in the implementation of the AFTA rules of origin can provide some lessons for the upcoming FTAs. Rules of Origin in AFTA: How Are They Working in Practice? AFTA preference utilization rates are low. ASEAN countries have implemented unilateral trade liberalization over the past two decades, and achieved low MFN rates (with average tariffs around 7 percent) by the time AFTA was implemented. This contributed to a limited impact of AFTA today, less than 5 percent of intra-asean trade makes use of the AFTA preferences. This is low compared with other FTAs. There are several reasons for the low utilization rates of AFTA: difficulty in satisfying the required value added requirement, difficulty in proving that the required value added has been satisfied, low preference margins, and high administrative costs of compliance. When the costs of complying with the rules of origin exceed the margin of preference then the trade agreement becomes irrelevant and trade will take place under the MFN regime. The value added rule is simple in principle, but difficult to comply with. The AFTA rule of origin of 40 percent value added is simple and transparent, but it is proving difficult to implement in practice. AFTA members, especially CLMV countries, are often unable to cumulate the i

8 necessary local/regional content. This is partly due to the high degree of production fragmentation in East Asia half of its trade is in electronics and machinery where production networks are widespread. The import content (from outside ASEAN) of export is high, making it difficult to comply with the 40 percent valued added rule. Further, countries with low labor costs will find it more difficult to comply with a given value-added requirement than higher labor cost countries. The value added rule is also vulnerable to exchange rate fluctuations any movement in the exchange rate leads to a change in import costs. This becomes problematic when the exchange rate fluctuations are widespread such as during the 1997 financial crisis. Moreover, the cost of proving origin is high. Computation of costs, invoicing, and other documentation demands inherent in the value added rule are complex, especially for smaller firms or firms from less developed East Asian countries. The administrative cost of compliance to prove origin is a deterrent for the use of preferences. The complexity of administering and complying with the ROO exacerbates the problem. Surveys in a range of ASEAN countries highlight concern over the time and paperwork involved in obtaining Form D (official form to prove origin in AFTA), and the large amount of documentation required to prove origin (including invoices and other evidence to each input used in the final product). These problems are particularly acute for small firms and for firms for whom prompt delivery is a key element of competitiveness. The requirement that all Form Ds should be issued by designated government departments significantly increases the compliance costs compared to many other FTAs where private sector associations are permitted to issue certificates of origin. Estimates of the costs of requesting preferences within AFTA might be in the range of percent larger than those of other preferential schemes. Moreover, customs valuations differ across countries, pre-export inspections required by AFTA add to cost, transactions remain time-intensive and required face-to-face contact with officials, and incoming goods enjoying preferences are randomly subjected to post-audit checks. Preference margins for products traded within ASEAN are low. Another reason for the low utilization rates of AFTA preferences is the low margin of preference on the products that are traded in large quantities within ASEAN, especially compared to the cost of compliance. Intra- ASEAN trade is dominated by computer/machinery and electrical equipment where the tariffs are very low (around 1.5 percent), making AFTA preferences largely irrelevant. Products with the highest margins of preference typically have a low value of import as a share of total intra- ASEAN trade such that the 40% value-added rule of origin is a binding constraint to preferential trade. This is known as the snow-plough effect in the AFTA agreement, vehicles especially designed for traveling in snow are given a high preference margin, but are irrelevant for ASEAN trade. Moreover, countries that confer the highest margins also appear to impose non-tariff measures on these same products (such as quantity control measures on certain categories of vehicles). The bulk of intra-asean trade occurs in commodities where preference margins are below the threshold that would justify the cost of compliance. Estimates based on other FTAs show that preferences start to have a trade stimulating effect only when preferential rates are at least 25 percentage points lower than the MFN rates. Over 90 percent of intra-asean4 (Malaysia, Indonesia, Thailand, Philippines) trade occurs in commodities where preferences are below 25 percent the threshold for using the preference. Only about 8 percent of eligible trade flows have a preferential margin above 25 percent (and are therefore worth using ). ii

9 Moving Forward: What Is Being Done, What Can Be Done? The value added rule could be relaxed to 30 percent especially for the less developed countries of ASEAN. ASEAN should stick to a simple and transparent ROO. Even though a value added rule is potentially costly for small firms in poor countries because of its requirements in terms of accounting, it has the advantage of being unambiguous and bypasses the need for product-specific rules, and does not leave leeway to lobbying by domestic industries over specification of rules. Analysis of data from the EU shows that a decline in the value-added requirement would tend to increase utilization rates. There a reduction in the maximum regional content from 60 percent to 50 percent is predicted to lead to an increase in the utilization rate by 2.5 percentage points (from the estimated mean of 17 percent) in the case of GSP and by 8.2 percentage points (from the estimated mean of 13 percent) in the case of ACP. Increasing utilization rates are likely to be most apparent for the low wage CLMV countries. The decision on whether to lower the value-added requirement should amount to an assessment of whether the potential gains in terms of greater regional trade significantly outweigh the risks of trade deflection. The analysis in these papers and the experience from elsewhere suggests that a 30% value-added requirement will be sufficient to prevent any significant trade deflection. A CTC rule could be introduced in parallel to not instead of a value added rule. In order to ease the potentially cumbersome procedures involved in the valuation and certification of declared costs under the RVC rule, ASEAN members are considering shifting to the change in tariff classification (CTC) criterion for all products. This would be a good option if it is introduced as an additional choice to (and not instead of) the value added rule, is implemented as a common rule across products (for example change at the heading or 4 digit level) and if additional product-specific exemptions to the general rule are kept to a minimum. There are risks in such an approach that will need to be suppressed, especially if protection groups seek to increase the number of exceptions to the general rules and influence the specification of these rules by excluding inputs from certain tariff headings, attaching an essential process in the transformation of the product, or turning the value added rule as an additional instead of an alternative rule to satisfy origin. Bilateral FTAs with restrictive product specific rules will also most likely define the parameters of wider FTAs formed later on, as in the case of the Japan- ASEAN EPA. The advantage of introducing change of tariff heading as an alternative rule is that it is a rule that is very easy for Customs authorities to understand and implement, since they are dealing with issues of customs classification in their every day work. In addition, proving compliance can be easy for small firms since the presentation of the import and export invoices showing different customs codes at the heading level is likely to be sufficient. Administrative procedures to obtain origin could be simplified. In order to ease delays and costs involved in proving compliance with ROO, self-certification can be introduced. This was proposed by Australia, Korea, China, and Japan in their respective FTA negotiations with ASEAN. For FTAs such as AFTA which group countries with a wide variation in MFN tariff levels, avoiding trade deflection will always be a difficult challenge given the complexities of verifying the origin of goods produced or assembled from multiple locations worldwide. The burden of proper verification becomes all the more taxing in an environment where the proliferation of bilateral FTAs leads to numerous potential backdoors that need effective policing. The likelihood of trade deflection increases when relatively high tariff countries like Thailand or the Philippines also acquire access to pockets of low barriers. This in turn might lead to more intensive or heavy-handed verification procedures that will further hike administrative and waiting costs. If implementation indeed becomes too difficult, leading to the inability to arrest a iii

10 significant amount of trade deflection, it may have a direct adverse consequence on the level of domestic political support for the FTA. The lack of coordination in setting ROOs amidst the proliferation of FTAs has a political cost attached to it, namely the cost of choosing favorites among favorites. Different permutations in the exchange of concessions among countries result in ROOs with varying degrees of restrictiveness, which in turn lead to a hierarchy of partners not unlike the EU s so-called pyramid of preferences. Differentiating partners into friends, lesser friends, and foes has bred all sorts of animosities, and has not created an environment conducive to the development of closer or strategic economic partnership. The recent surge of FTAs in East Asia may be less about trade and more about issues like trade facilitation or regulatory barriers involving investments and services, where negotiations have bogged down in the multilateral arena. It is not unlikely that irritations stemming from contentious ROO negotiations, or the uneven restrictions applied to trade among different partners, could spillover to more important, high-stake negotiation areas. Key lessons for future agreements include: -Simple and transparent rules -Give producers choice of methods to satisfy origin -Ensure minimal costs of documentation -Ensure minimal costs of implementation for Customs perhaps raise spectre of customs have to deal with a multitude of different product specific rules of origin, different documentation etc - Issue of cumulation in ASEAN but not in bilaterals being signed by ASEAN members, as this would increase the risks of hub and spoke with US, EU, Japan and maybe China. iv

11 PART I RULES OF ORIGIN IN EAST-ASIA AND LESSONS FROM OTHER REGIONS

12

13 Chapter 1 Rules of Origin in East Asia Robert Kirk, The Services Group Inc. The consolidation of the multilateral trading system following the establishment of the World Trade Organization (WTO) in 1994 has been accompanied by the rise of regional preferential trade agreements (RTA). The Sutherland Report (2005) observes that the rise of RTAs has virtually reached the point where the core principle of non-discrimination embodied by the mostfavored-nation (MFN) clause and the national treatment provisions which is integral to the WTO is now exceptional treatment or least favored treatment. With more than half of total world trade taking place under RTAs and rising, and with new agreements being implemented, more attention is being focused on the impact of preferential rules of origin (ROO). ROO are now center stage in determining market access under RTAs. ROO are a necessary feature of any form of RTA since they are required to ensure the preferences are available only to the signatories to the agreement. The sheer number of RTAs has resulted in a complex web of multiple preferential tariff rates being applied to various trading partners, frequently with varying timelines for reducing tariffs. Complex and inconsistent ROO which often govern eligibility criteria of RTAs have the potential to restrict effective market access and can prevent many of the RTA benefits from being realized. Recent work has focused on unscrambling the impact of rules of origin on economic welfare with studies on NAFTA, Latin America, and Southern Africa showing how complex preferential ROO result in increased transaction costs and limit the use of the trade preferences. 1 With the gradual accumulation of applied studies there is increasing evidence that the preferential ROO will reflect the specific protectionist interests of the more powerful partners in the preferential agreement (Moise, 2003). Until recently regionalism in East Asia was limited to the ASEAN Free Trade Area (AFTA) which accounted for a small proportion of total trade. Since 2000, however, there has been an upsurge in new regional agreements with ASEAN and the larger ASEAN economies of Thailand, Singapore, Indonesia and Malaysia. Despite the AFTA and the rise of regional agreements the majority of trade in East Asia continues to take place on an MFN basis. This paper assesses the various ROO; examines the ROO within AFTA and their potential impact on preference utilization; and presents policy options for improving the ROO. 1 Preferential Trade Agreements and the Pattern of Trade in East Asia In the mid-1990s economies in East Asia reduced tariffs unilaterally and sought to promote investment through the development of export processing zones, the widespread use of export duty rebates, and sub-regional cross border arrangements such as growth triangles. These incentives ensured that many exporters were entitled to import at preferential or duty free rates without using the evolving AFTA. Following the Asian financial crisis of East Asian economies began to explore ways of increasing policy coordination and cooperation in trade and investment as a way of reducing the risks of financial contagion. At the same time there was a growing awareness within ASEAN of the importance of engaging more with China. Further the ASEAN private sector began to lobby for preferential access to regional markets. These internal pressures for increasing regional preferences have been buttressed by concern at the lack of substantive progress of the Doha 1 The Origin of Goods, Cadot (2006); OECD, (2003); European Commission, (2003). 3

14 Round and the absence of momentum in the Asian Pacific Economic Cooperation (APEC) to advance the trade agenda. By the end of the 1990s many East Asian economies considered the negotiation of regional and extra-regional FTAs to be in their strategic interests. Table 1: Major RTAs including countries in East Asia Partners Status Year Intra-Regional Inter- Regional Bilateral RTAs Singapore-Australia Implemented 2003 X Singapore-New Zealand Implemented 2001 X Singapore-Japan EPA Implemented 2002 X Singapore-EFTA Implemented 2003 X Singapore-US Implemented 2004 X Korea-Chile Implemented 2004 X Japan-Mexico Entered into force 2005 X China-Thailand (agric.) Entered into force 2003 X China-Hong Kong Entered into force 2004 X China-Macao Entered into force 2004 X Korea-Chile Entered into force 2004 X Thailand-India TIG Entered into force 2004 X Thailand-Australia Entered into force 2005 X China-Chile Entered into force 2005 X Thailand-New Zealand Entered into force 2005 X China-Pakistan Entered into force 2006 X Singapore-Korea Signed 2005 X Korea-US Signed 2007 X Japan-Thailand Signed 2007 X Japan-Malaysia Signed 2005 X Japan-Philippines Signed 2006 X Singapore-Jordan Signed 2004 X Singapore-Mexico Under Negotiation since 1999 X Singapore-Canada Under Negotiation since 2001 X Korea-Japan Under Negotiation since 1998 X Japan-Indonesia Under Negotiation since 2005 X Japan-Vietnam Under Negotiation since 2006 X Thailand-US Under Negotiation since 2004 X Thailand-EFTA Under Negotiation since 2005 X Thailand-India Under Negotiation since 2004 X Thailand-Peru Under Negotiation since 2004 X Thailand-Bahrain Signed FA 2002 X Korea-New Zealand Joint Study 2000 X Korea-Thailand Joint Study 2001 X Japan-Chile Joint Study 2000 X Korea-Mexico Discussion (halted) 2000 X Korea-Australia Discussion 2000 X Japan-Canada Discussion 2000 X Plurilateral FTAs AFTA Implemented 1992 X Chile-NZ-SGP-Brunei Entered into force 2000 X ASEAN-Korea TIG Signed 2002 X ASEAN-India Under Negotiation since 2003 X ASEAN-Australia Under Negotiation since 2004 X ASEAN-China Signed FA 2001 X ASEAN-Japan Framework Agreement 2002 X China-Japan-Korea Joint Study 2001 X ASEAN-CER Ministerial Declaration 1999 X ASEAN-3 Discussion 2000 X Source: Agreements and 4

15 By 2007, Asian countries (excluding China) had ratified 13 bilateral and regional FTAs and had negotiated (but not implemented) another seven (Table 1). Many more bilateral and regional FTAs are currently being actively negotiated by Asian countries many of these are with non- Asian partners. Thailand entered into Bilateral Trade Agreements with Australia and New Zealand in 2004 and 2005 respectively. It has recently concluded negotiations with Japan and is currently engaged in negotiations with Bahrain, EFTA, India, Peru and the US. The Philippines signed an agreement with Japan in 2006, while Korea, Indonesia, Vietnam, Laos and Cambodia are currently still negotiating with the latter. Over the past six years China has negotiated an FTA with ASEAN and has completed five bilateral FTAs with Thailand in 2003 (for agricultural produce), Hong Kong and Macao in 2004, Chile in 2004 and Pakistan in China is currently negotiating or initiating another 17 bilateral and regional FTAs. ASEAN has further signed initial agreements with Australia and New Zealand, and with Japan. It is currently negotiating with India and South Korea (a Trade in Goods Agreement between Korea and ASEAN is already in place). Of the 18 agreements currently in effect, 7 involve only East Asian countries and the remaining 11 are bilateral or trilateral agreements between an East Asian country and a country outside the region. AFTA. The ASEAN Free Trade Area was established in 1992 between the six ASEAN member countries Indonesia, Malaysia, the Philippines, Singapore, Thailand and Brunei with the objective of increasing their international competitiveness by lowering intra-regional tariffs. Through the Common Effective Preferential Tariff (CEPT) scheme, tariffs would be reduced to 0-5 percent within a 15-year period. By 2003, after only 10 years, AFTA succeeded in reducing average tariff rates from 11.4 percent in 1993 to 2.4 percent for the original ASEAN-6. The new ASEAN members Cambodia, Laos, Myanmar and Vietnam (CLMV) joined AFTA in the latter half of the 1990s and committed to removing tariffs over a longer time period. The AFTA approach to free trade is complex with countries only receiving preferences for items on both the counterparty s Inclusion List as well as items on their own Inclusion List. Consequently all products on both offering countries (and offers are made bilaterally and not to all ASEAN members) in the Temporary Exclusion, General Exception and Sensitive Lists are not offered tariff preferences. About 95 percent of tariff lines are in the Inclusion List. But this fractured approach means that the degree of market access faced by an AFTA exporter of any specific product could vary according to the specific ASEAN market. ASEAN has agreed to systemic rules of origin, product exclusion practices and phase-in modalities. Table 2: AFTA CEPT rates and distribution of tariff lines 2003 Country CEPT Inclusion Temporary General Sensitive List List Exclusion List Exception List Brunei Darussalam Indonesia Malaysia Philippines Singapore Thailand ASEAN Cambodia Lao PDR Myanmar Vietnam New Members(CLMV) ASEAN Source: ASEAN Secretariat 5

16 ASEAN+China. In November 2000 ASEAN and China agreed to establish a free trade area and concluded a Framework Agreement in This committed both parties to establish an FTA by Two way ASEAN China trade had increased rapidly and by 2000 reached $39.5 billion to account for 8.3 percent of China s foreign merchandise trade. ASEAN is now China s fifth largest trading partner and a major supplier of raw materials, as well as textiles, metal and metal products, leather and leather products, chemicals, electronic components and plastics. Bilateral trade with China represents just 3 percent of total ASEAN exports and 5 percent of ASEAN imports (based on trade with the ASEAN 6). Under the ASEAN-China agreement, China will establish an FTA with the original ASEAN 6 by 2010 and with the newer ASEAN members by Like AFTA, the ASEAN-China FTA is complex with each party selecting their own sensitive list and bilateral market access depends upon the interaction of the offers of the two countries. The ASEAN FTA with China is effectively 10 FTAs as each member of ASEAN submits a country specific offer to China which then prepares an offer for each country. Preferences are offered on a reciprocal basis so will only apply to those specific products that are included in both countries lists. The rules of origin are identical to the AFTA ROO although there is a provision for these to be negotiated between the parties. The Early Harvest program, which cuts tariffs on eight categories of agricultural products (about 600 different tariff lines), also uses the existing ASEAN ROO. The Early Harvest tariff reductions to zero were to be in place for the original ASEAN members by 2006 and for the CLMV by The Philippines elected not to participate in the Early Harvest Program arguing that their agricultural sector would be adversely affected. Two-way trade between China and ASEAN increased by 40 percent in 2003 and this trend has continued. 2 AFTA Utilization of Preferences ASEAN countries have implemented a unilateral trade liberalization over the past two decades and achieved low MFN tariff rates by the time AFTA was implemented (Figure 1). This contributed to the limited impact of AFTA although AFTA implementation is measured in terms of the number of tariff lines that have been either partially or fully liberalized, less than 5 percent of total intra-asean trade takes place under the CEPT. Thus 95 percent of intra-asean trade found it more beneficial to either pay the MFN tariff or take advantage of other schemes such as the duty drawback programs or duty-free treatment in the EPZs. At the country level, only 11 percent of Thailand s imports from ASEAN took advantage of the CEPT while only 4 percent of Malaysia s AFTA exports benefited from preferences in The reasons for the low utilization rates are straightforward: AFTA s margin of preference on the products that are traded in large quantities are too small to compensate for the administrative costs of applying for preferential tariff treatment. Intra-ASEAN trade is dominated by computer/machinery (HS chapter 84) and electrical equipment (HS chapter 85) where the tariffs are very low 1.5 percent and 1.4 percent, respectively and AFTA preferences not relevant (Figure 2). Recent work on both the EU and NAFTA indicates that when the preferential margin drops below 5 percent traders choose to use the MFN because the compliance costs tend to exceed the preference. 2 In the case of AFTA the margins of preference are below 5 percent for most ASEAN countries, and thus traders find it more advantageous to use the MFN tariff. 2 See Cadot et al (2005) which estimates the compliance costs associated with the rules of origin under NAFTA and the PANEURO at 6.8 and 8 percent of the value of the product respectively. 6

17 Figure 1: Average tariff rates in East Asian countries Figure 2: AFTA margin of preference (difference between MFN rate and CEPT rate) MFN tariff (%) Automotive China Japan Korea Singapore Malaysia Thailand Indonesia Philippines Textile and Garment s Elect ronics 5 Overall margin of pref erence, percent Source: ASEAN Secretariat Source: UN-COMTRADE, 2004 The complex sourcing of inputs from across South East Asia and elsewhere to produce a finished product for export to the region and major developed markets requires an efficient trading system where components and intermediates can be sourced at world price (or close to world prices) with timely delivery and low trade facilitation costs. Transnational corporations will be continually reviewing their sourcing policies to purchase from the most efficient suppliers. To date this has developed independently of the AFTA and the ROO governing the CEPT. However, ongoing trade negotiations may result in changes to the ROO that have implications for the sourcing of inputs by Factory Asia. 3 Preferential Trade Agreements and Rules of Origin Rules of origin are necessary within a preferential trade agreement to prevent trade deflection or trans-shipment, which occurs when goods enter through a low duty country and then are shipped to a higher duty country. Preferential ROO aim to ensure that only goods originating in the participating countries (i.e. the members) receive the preferences. Products that meet the ROO criteria are referred to as originating in a member country and will generally qualify for the preferential tariff rates. The standard procedures for administering ROO are set out in the Kyoto Convention (1977) on the Simplification and Harmonization of Customs Procedures. This states that ROO should, inter alia, be objective, understandable and predictable, and be capable of being administered in a consistent, uniform, impartial and reasonable manner. ROO can be designed to achieve a specific commercial policy objective, such as protecting existing regional producers of intermediate or final products. The impact of ROO on production costs stems from the technical criteria imposed by the ROO regime. If the ROO result in an increase in the cost of intermediate goods from the pre-rta levels there is the potential for trade diversion in intermediates within the RTA. There are two broad types of ROO. Either a good consists entirely of inputs from member countries in which case it is deemed wholly obtained or produced and qualifies, or the product uses a combination of member inputs and imported inputs. In the latter case it is necessary for the product to be substantially transformed in order to qualify. The way in which substantial 7

18 transformation is defined determines the economic impact of the ROO and in turn the relevance of the RTA to the members. Substantial transformation can be defined in a simple transparent manner in order to meet the authentication requirement that the product has been finished in the originating country. It is possible to list a small number of rudimentary screwdriver activities that are deemed as not eligible in order to prevent circumvention. This generally covers activities such as mixing, cleaning, washing, and repackaging. However, when ROO accompanying many RTAs are examined, it is clear that the trade negotiators eschewed a simple approach in favor of crafting more arcane requirements. In many RTAs the ROO run to several hundred pages and in some well-known cases one product can require up to twenty pages of description. 3 Even relatively simple products can have astonishingly complex ROO. 4 A priori, a simple ROO is not necessarily less restrictive than a more complex ROO. Practical approaches to ROO. There is no unique approach to defining originating criteria. In practice between one and three approaches can be used either on their own or in any combination with each other: Change in Tariff Classification (CTC) between the manufactured good and the intermediate inputs used in the process from countries outside the RTA. The CTC may require the product to alter its chapter (two digits at the HS level), heading (four digits), sub-heading (6 digits) or item (8-10 digits) in the exporting country. The ROO may attach Exceptions (EX-out) to a particular CTC these generally prohibit the use of nonoriginating materials from a specified subheading, heading, or chapter; Value Content (VC) requires the product to acquire a certain minimum local value in the export country. The value content is expressed in one of three ways-as the minimum percentage of value that must have been added in the exporting country (domestic or regional value content, RVC), as the difference between the value of the final good 5 and the costs of the imported inputs (imported content), or as the value of the parts where originating status is given to products meeting a specified minimum percentage of originating parts as a proportion of the total; Technical Requirements specify the production processes that must take place or prescribes or prohibits the use of certain inputs. This is widely used for apparel producers. In addition to the product-specific requirements for substantial transformation each RTA ROO typically contains general rules. These may include: 3 The NAFTA product specific ROO exceed 300 pages of text. For a sector example refer to the ROO for motor vehicles under the NAFTA Agreement For a summary of the negotiations on motor vehicles under the NAFTA refer to Negotiating NAFTA, Robert Maryse (2000) 4 Brenton and Imagawa (2004) cite the case of the EU ROO for fish under the Cotonou Agreement. This requires that the fish be caught in the territorial waters of an eligible (ACP) country, and when landed at an EU port the ship should carry documentation establishing that the vessel s captain, officers and at least 50 percent of its crew were nationals of an EU or ACP state, the ship is registered in an EU or ACP state, it sailed under the flag of an EU or ACP state, it was at least 50 percent owned by nationals of an EU or ACP state (although there are certain exceptions for leased and chartered vessels), and the chairman and the majority of the board members of the company owning the vessel were nationals of an EU or ACP state. 5 It is important to note whether the price refers to cost or production or ex factory price (i.e. the sale price) 8

19 De minimis allows for a specified maximum percentage of non-originating materials to be used as though they were originating. This provides some flexibility in the CTC and Technical Requirements by making it easier to use non-originating products; The roll-up or absorption principle allows for initially non-originating inputs that have acquired origin through meeting specific processing requirements to be deemed as originating when they are used as inputs in a subsequent transformation; Cumulation allows for producers in one RTA country to use materials from another RTA country (or countries) without losing the preferential status on the final product. This concept is particularly important when considering the impact of multiple RTAs. There are several different types of cumulation. Bilateral cumulation functions between two RTA partners and permits them to use products that originate in the other RTA partner as if they were their own when seeking to meet the ROO. Under diagonal cumulation, countries in the same RTA can use products that originate in any part of the common ROO zone as though they had originated in the exporting country. Full cumulation is more flexible than diagonal cumulation as it allows countries tied by the same ROO regime to split the intermediate processing between the parties to the RTA and be granted the preference as long as when all the materials and processing used are added they are sufficient to meet the ROO; A separate list of activities or processes which are considered insufficient to confer origin; this may include cleaning, sorting, marking, labeling; Explicit prohibition on the use of duty drawback which rules out the refunding of tariffs on non-originating inputs that are used as inputs into the final product. Many countries use duty drawback or duty rebates as an investment incentive for exporters; however, in the case of a RTA this is considered to provide a cost advantage to the RTA-based producers for export compared to the producers for the domestic market. Many duty rebate schemes require the producers to export a minimum of 80 or 90 percent of their total sales. The procedures for certifying the origin of the goods can potentially impose high administrative costs on exporters. Common certification methods are self-certification by exporters, certification by the exporting country government or an industry umbrella group that has been accorded responsibility for issuing the certificate. Cumbersome and time-consuming certification requirements increase the costs of compliance and reduce the incentives to use the RTA. ROO vs. local content schemes what is the difference? ROO are used as instruments of trade policy with trade negotiators expending substantial time and energy carefully crafting both the general rules and the specific criteria for substantial transformation. ROO may be considered as analogous to content protection or local content schemes. These schemes may be considered as a type of proportionally distributed quota in which duty-free (or at preferential rates) imports are permitted in some specified proportion to purchases from domestic producers. The local content scheme requires that domestic final good producers purchase a specified minimum proportion of their intermediate goods from domestic firms. ROO extend this to apply to sourcing either locally or from member countries of the RTA. The literature on local content schemes has shown convincingly the harmful economic impact. In fact, ROO are more complex than local content schemes since firms are permitted to source from within the RTA and a priori it is impossible to define if the ROO are restrictive or not. Any ROO that results in producers and investors modifying their sourcing behavior compared to what would prevail without the RTA is restrictive. This may apply to ROO that are simple. Indeed a ROO that is binding for one country may be irrelevant for another member of the RTA. For example, a country with substantial resources and a diverse manufacturing base will ceteris 9

20 paribus be able to comply with ROO relatively more easily than a small economy with a narrow specialized manufacturing sector. Measuring the restrictiveness impact of ROO. In order to assess the degree of restrictiveness of ROO, Estevadeordal and Suominen (2006) develop a facilitation index, which seeks to summarize information on regime-wide rules. Quantifying non-tariff barriers is fraught with difficulties but when conceptualized in terms of tariff equivalents it becomes easier to assess their economic impact. ROO are a particularly difficult form of non-tariff barrier to classify because their impact cannot be known ex ante. The approach taken by Estevadeordal is to classify ROO according to their administrative types, make prior assumptions about the restrictiveness of each type, and assign an index between 1 and 7 in order of increasing restrictiveness of each type (Table 3). There are three main types of ROO: change of tariff heading (CTH) or change of tariff classification (CTC), value content and technical requirements. These are not mutually exclusive with some products requiring a combination of the different types. Table 3: Index of restrictiveness of rules of origin Index of Restrictiveness Description 1 (least restrictive) CTC at the level of HS 8-10 digit 2 CTC at the level of subheading HS 6 digit 3 CTC at subheading plus value content 4 CTC at the level of heading HS 4 digit 5 CTC at the level of heading and value content 6 CTC at the level of the Chapter HS 2 digit 7 (most restrictive) CTC at the level of the Chapter and Technical Requirement Source: Estevadeordal et al in Cadot (eds.) While the restrictiveness assumptions made by Estevadeordal and Suominen to come up with the restrictiveness index are plausible, a problem which they note is that the HS coding system was not designed as a policy implementation tool for ROO. As experience has shown in a number of RTAs, a change of chapter requirement in most agricultural or primary commodities might be much easier to meet that even a change in subheading in some chemical, electrical or mechanical goods sectors where inputs and outputs are often included in the same subheading. While changes in tariff chapters, subheadings and headings may seem to be relatively straightforward, how should technical and value requirements be categorized? Is a technical requirement more or less restrictive than a restriction on import content? And if there is both a technical and value-content requirement in a sector, is that more restrictive than only having a technical requirement or a stand-alone value requirement? The answer cannot be known ex ante and will depend at least in part on the nature of the technical requirement and on the amount of import content allowed under the value requirement. When there are two requirements, the answer depends on which one is binding for producers or potential regional exporters in the sector. The restrictiveness of any given requirement will vary according to the sector in which it is applied. Yet, in the index a 60 percent local content requirement receives the same restrictiveness index as a 40 percent requirement and almost certainly the impact will be very different. Great caution is needed in interpreting the restrictiveness impact of different ROO across sectors. Uniform ROO across all sectors will impact those sectors asymmetrically. ROO that are different across sectors may not easily show the extent of their restrictiveness ROO that show many sectors with extremely liberal ROO may not indicate an increase in market access under the RTA; in such cases the strict ROO may only apply to few sectors, but these may be the protected ones with potential trade creation. The variation in the ROO may be evidence of made 10

21 to measure protection aimed at ensuring the partner countries are unable to compete with the existing import substituting sector. 6 4 The ASEAN Rules of Origin The ROO for the ASEAN Free Trade Area (AFTA) originally applied a single method for all products, namely value added. Under AFTA, originating status is conferred under either one of two conditions: (a) products wholly produced or obtained in the exporting Member States as defined in Rule 2, 7 or (b) products not wholly produced or obtained: (i) A product shall be deemed to be originating from ASEAN Member States, if at least 40 percent of its content originates from any Member States. (ii) Locally-procured materials produced by established licensed manufacturers, in compliance with domestic regulations, will be deemed to have fulfilled the CEPT origin requirement; locally-procured materials from other sources will be subjected to the CEPT test for the purpose of origin determination. (iii) Subject to sub-paragraph (i) above, for the purposes of implementing the provisions of Rule 1 (b), products worked on and processed as a result of which the total value of the materials, parts or produce originating from non-asean countries or of undetermined origin does not exceed 60 percent of the FOB value of the product produced or obtained and the final process of manufacture is performed within the territory of the exporting Member State. The ASEAN ROO provide for full cumulation where the imported input is wholly originating and partial cumulation in all other cases subject to a minimum threshold of 20 percent. In cases where the imported input does not meet the ROO (i.e., less than 40 percent local/asean content) but the local/asean content exceeds 20 percent, the intermediate input will not receive the CEPT preference but must complete a Certificate of Origin (referred to as a Form D) in order to be used for cumulation. These implementation guidelines were agreed at the ASEAN Economic Ministers Retreat in April Provision is made for products to transit through a non-asean country subject to it being justified for geographical reason or by consideration related exclusively to transport requirements. The ASEAN ROO does not mention the treatment of duty drawback or the Absorption or Roll Back principle. The AFTA does not require a Form D for small scale cross border trade, defined as less than $ This was the case in SADC where South Africa negotiated more stringent ROO for those sectors in which had a commercial interest (see Erasmus et al, 2006). 7 Rule 2 includes the definition of wholly produced products within a Member country which includes mineral products extracted from its soil, products obtained by hunting or fishing there, products of sea fishing and other marine products taken from the sea by its vessels, agricultural products harvested there. 11

22 To qualify for the CEPT the products require a Certificate of Origin issued by a government authority designated by the exporting Member State and notified to the other Member States in accordance with the Certification Procedures that are approved by the Senior Economic Officials Meeting (SEOM). ASEAN Member States have approved Operational Certification procedures for the issuance and verification of the Certificate of Origin (i.e. Form D). Form D can only be issued by Government authorities, unlike many FTAs where certificates of origin can be issued by private sector agencies such as Chambers of Commerce. In order to receive a Form D the exporter must apply in writing to the relevant government Box 1: Obtaining a Form D in Vietnam In Vietnam, the Export-Import Managing Department of the Ministry of Trade is the issuing institution for Form D. An application is submitted to an inspection company authorized by the Ministry of Science to conduct a cost screening to ensure local content of 40 percent or more. VINACONTROL remains the largest inspection firm, but the number of authorized companies has increased over the past few years. This provides for competition. Screening generally takes between one-half to a full day. The applications required for each shipment are submitted to a branch office of the Export-Import Managing Department (9 Branches nationwide) and are accompanied by a certifying letter from the inspection company, a commercial invoice, a Customs declaration form, a bill of lading, and a copy of the exporter s commercial license. Form D is issued within 2 hours. authority requesting pre-export verification of the origin of the products (see Box 1 for the Vietnam example). In the case of locally procured materials (i.e. Rule 1) the final manufacturer is entitled to complete a self-declaration. Rule 6 of the operational guidelines requires the Government authorities to carry out proper examination upon each application for the Certificate of Origin. The operational guidelines are very thorough and appear to have been drafted to facilitate the use of ROO. The AFTA ROO are simple but implementation shows that meeting a 40 percent value added is difficult, especially for CLMV countries. The basic rules of origin adopted by ASEAN for the AFTA are simple and transparent, with a requirement that Members must source at least 40 percent of the value of the ex factory price from ASEAN. However, once Members began implementing this basic ROO, concerns were raised about the difficulty of meeting the 40 percent local content. This ROO was especially difficult for the CLMV countries to meet outside of agricultural commodities. Data on aggregate value added as a percentage of total output (in domestic prices) across sectors and countries show very few sectors with a value added above 40 percent (Table 4). Table 4: Value added as proportion of total costs (in percent) Industrial Sector Cambodia Indonesia Laos Philippines Thailand Vietnam Processed Meat/Fish/Vegetables Wearing Apparel Leather Processing Wood Processing Paper and Paper Products Basic Chemicals Rubber Products Glass Products Basic Iron and Steel Products Metal Products Electrical Motors Motor Vehicles/Auto Furniture Note: Laos and Philippines 1999, Cambodia, Thailand and Vietnam 2000 and Indonesia 2002 Source: Derived from Value Added Surveys from UNIDO. 12